Djoerd Hiemstra’s home page

The DB Colloquium of Tuesday 29 January, 14:00 h.-15:00 h. in ZI-3126 consists of two small presentations.

Comprehending historical election programs using XML and XRPC

by Douwe van der Meij

Party programs for elections can be incomprehensible, not to mention the comparison of current party programs to that of a decade ago. This paper focusses on a way to compre- hend the latter. It shows how to use xml to store election programs and to query those. This paper also comes with a proof of concept (PoC). In retrospect we look at this PoC, and we discuss the design choices made.

---

Boeken zonder leeftijdscategorie sneller vinden

by Wout Maaskant

In dit onderzoek is een systeem ontwikkeld waarmee gebruikers sneller boeken waar geen leeftijdscategorie aan is toegekend kunnen vinden in een bol.com-corpus, door gebruik te maken van eigenschappen van vergelijkbare boeken waar wel een leeftijdscategorie aan is toegekend. De gelijkenis tussen boeken wordt bepaald met behulp van het vector space model.

Storing Personal Information Management (PIM) data is not trivial, because of the variety in content types. Existing PIM storage systems have shortcomings in performance, data concistency and/or concurrency. In this thesis, we propose several optimisations and test them in Akonadi, KDE’s new central PIM data access manager. The optimisations include using the D-Bus protocol for transmitting short commands and notifications and an IMAP-compatible protocol for data access and modification. The PIM data is kept in its native format, but compressed and split up into separate, frequently-used parts for increased performance. Both the synthetic and use case based evaluation results show that the proposed modifications perform well and help maintain data consistency in Akonadi.

Read more on E-prints

We have two excellent presentations at the DB master students colloquium of Friday 30 November:

“Storing Personal Information Magagement Data”
by Robert Zwerus

“Cooperative Intelligent Transport Systems”
by Bobby Nijssen

When: Friday 30 November, 13.45 h. - 15.30 h. Where: ZI-3126

KDE, The K Desktop Environment: Conquer your desktop

Who: Volker Krause of KDE.org
When: Wednesday, 28 November 2007, 14.30 h. - 15.15 h.
Where: ZI-4126

Volker Krause of KDE.org will give an overview of what’s new in KDE4. He will talk about Akonadi, the Personal Information Management (PIM) Storage Service of KDE. Furthermore, Volker will talk about the currently ongoing cooperations between various universities and KDE (students working on KDE in practical courses, thesis on KDE topics, EU-funded research projects).

Time and Place of next week’s lecture on 17 October is changed to:

WA-204, at 10.40h. - 12.25h.

I would like to ask everyone to be present at the guest lecture. Martin Engels comes all the way from Leiden to talk about GIS. If you are unable to come, please let me know as soon as possile.

Title: GIS and Geodatabases: the ESRI approach
Speaker: Martin Engels, ESRI Netherlands
Room: WA-204
When: Wednesday 17 October 2007, 10:40 h. - 12:25 h.

About ESRI:
ESRI designs and develops the world’s leading geographic information system (GIS) technology. GIS is an important tool - one that helps shape the world around us. GIS technology helps fight forest fires, determine new national boundaries during peace negotiations, find promising sites for fast-growing companies, rebuild cities around the world, support optimal land-use planning, route emergency vehicles, monitor rain forest depletion, contain oil spills, and perform countless other vital tasks every day. Today, ESRI has more than 4,000 skilled employees worldwide who work with hundreds of business partners and tens of thousands of users.

Title: The evaluation of information retrieval systems by Wessel Kraaij (TNO-ICT) Wednesday 4 October, 13.45 hour in LA-1812 This lecture provides the tools and methodology for comparing the effectiveness of two or more information retrieval systems in a meaningful way. Several aspects of information retrieval systems can be evaluated without consulting the potential users or customers of the system, such as for instance the query processing time (measured for instance in miliseconds per query) or the query throughput (measured for instance as the number of queries per second). This lecture, however, focuses on aspects of the system that in uence the quality of the retrieved results. In order to measure the quality of search results, one must at some point consult the potential user of the system. For, what are the correct results for the query “black jaguar”? Cars, or cats? Ultimately, the user has to decide….

We are going to start a new colloquium on information retrieval related topics. It should bring together people working in this field (from different floors of this building) to discuss their newest research as well as new developments in IR in general.

For the first session, that will be held on Tuesday 26.9. 11:00 in our meeting room No. 3126, Claudia has volunteered to give a report on the Clef conference in Alicante (just 2 days after she returns from Spain. So you will get the newest information possible). We will further discuss how we are going to continue these colloquium meetings.

So if you are interested in IR topics, put this date in you agenda…

The Potential of User Feedback through the Iterative Refining of Queries in an Image Retrieval System

by Maher Ben Moussa and Marco Pasch

Inaccurate or ambiguous expressions in queries lead to poor results in information retrieval. We assume that iterative user feedback can improve the quality of queries. To this end we developed a system for image retrieval that utilizes user feedback to refine the user’s search query. This is done by a graphical user interface that returns categories of images and requires the user to choose between them in order to improve the initial query in terms of accuracy and unambiguousness. A user test showed that, although there was no improvement in search time or required search restarts, iterative user feedback can indeed improve the performance of an image retrieval system in terms of user satisfaction.

Maher and Marco are M.Sc. students at the University of Twente. They will present their work at the fourth International Workshop on Adaptive Multimedia Retrieval (AMR) in Geneva, Switzerland.

[download pdf]

DB master students colloquium, Friday 30 June at 13:30 h. - 14:30 h. in L-200

The following people will present (please note that room and time are not the same as usual):

• Yelei Zhang: Index processing for complex event detection
• Gert-Jan Poulisse: Compression and preprocessing of mammograms

At 15.00 that day there will be Design Project presentations and there will be drinks at about 16.00 to close the year. Hope to see all of you there!

Design project presentation, Monday 26 June at 14:30 h - 15.15 h. ZI-3126
by Sander van Loon, Bastiaan van den Berg, Vincent Groenhuis, Frank Zijlstra, Andre van Herk.

Abstract (in Dutch): Het doel van de LEGO 3DORopdracht was, een (LEGO) systeem te bouwen waarmee willekeurige objecten ingescand kunnen worden en er een driedimensionale, digitale representatie te maken. Door een LEGO Mindstorms systeem te gebruiken, in combinatie met visualisatieapparatuur (een camera, infrarood/afstandssensor, voelspriet, etc.) moest een driedimensionaal model worden opgebouwd dat een computer kan weergeven (bij voorkeur het systeem dat de scan heeft uitgevoerd). In deze presentatie wordt het ontwerp van het systeem toegelicht. Aansluitend om 15.45 u. is er een demonstratie van het systeem tijdens de vijfde Twente Student Conference in ZI-Oostzaal.